issues
search
HPDL-Group
/
Merak
Apache License 2.0
68
stars
9
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
Runtime error when trying to use the run_gpt language modelling example
#12
prajwal1210
opened
4 months ago
2
Support for UNet models?
#11
prajwal1210
opened
5 months ago
1
Merak is better?
#10
Lvjinhong
closed
7 months ago
1
Understanding Data Propagation and Communication in Model Parallelism
#9
Hongjie1Chu
closed
7 months ago
0
模型并行中不同层之间是如何通信的?同一层的张量并行组是如何通信的?
#8
Hongjie1Chu
closed
7 months ago
0
value error: func_inputs[k] = v
#7
huils20
opened
1 year ago
1
Introspecting pipeline stage partitioning results
#6
jaywonchung
opened
1 year ago
6
[per_device_train_batch_size] argument cause misunderstanding
#5
lin88lin8850
closed
1 year ago
2
self._valid_micro_batch(micro_batch_id) AssertionError when use "shifted_critical_path" train schedule
#4
lin88lin8850
closed
1 year ago
5
c10::CUDAError happens Occasionally
#3
lin88lin8850
opened
1 year ago
3
[fp16 error] In GPT-2 run example
#2
lin88lin8850
closed
1 year ago
2
RuntimeError: Tried to erase Node attention_mask_1 but it still had 1 users in the graph: {_assert_is_none: None}!
#1
QiaolingChen00
closed
1 year ago
4